Method and system for optimally allocating a network service

• US 7,502,857 B2
• Filed: 12/15/2000
• Issued: 03/10/2009
• Est. Priority Date: 12/15/2000
• Status: Expired due to Fees

• Current Assignee: International Business Machines Corporation
• Original Assignee: International Business Machines Corporation
• Inventors: Chen, Xiaoping, Morrissey, Christopher M., Barillaud, Franck
• Primary Examiner(s): Etienne; Ario
• Assistant Examiner(s): Chea; Philip J
• Application Number:
  US09/738,371
• Publication Number:
  US 20020075810A1
• Time in Patent Office: 3,007 Days
• Field of Search: 709/200, 709/238, 709/241, 709/221, 709223-224, 709/226, 709/205, 709/232, 709/235, 370/252, 370/254, 707/10, 718/105, 700/28, 715/734
• US Class Current: 709/226
• CPC Class Codes: H04L 43/0864 Round trip delays
• CPC Class Codes: H04L 43/12 Network monitoring probes
• CPC Class Codes: H04L 67/10 in which an application is distributed across nodes in the network software deployment G06F8/60; multiprogramming arrangements G06F9/46
• CPC Class Codes: H04L 67/1001 for accessing one among a plurality of replicated servers
• CPC Class Codes: H04L 67/10015 Access to distributed or replicated servers, e.g. using brokers
• CPC Class Codes: H04L 67/101 based on network conditions
• CPC Class Codes: H04L 69/329 in the application layer [OSI layer 7]
• CPC Class Codes: H04L 9/40 Network security protocols

Abstract

A network service allocation device including a probe, a module, and an engine. The device collects performance data relating to the physical characteristics of a distributed data processing system. The module utilizes the performance data to identify two or more node clusters on the distributed data processing system. The engine correlates the node clusters and one or more performance rules as related to the network service to provide a map of node cluster(s) for supporting the network service, or a list of one or more servers within a node cluster for supporting the network service.

Claims

1. A method for allocating a service in a distributed data processing system, said method comprising:

collecting a set of performance data representative of a set of physical characteristics of the distributed data processing system to form a collection of a set of performance data;

providing, using said collection of a set of performance data, cluster data that identifies a plurality of node clusters in said distributed data processing system, wherein a node cluster comprises an aggregation of nodes;

correlating at least one property of each of the of the identified plurality of node clusters with at least one performance rule required for supporting the service to determine a compliance of each of the plurality of node clusters to the at least one performance rule for supporting the service;

showing each node cluster of the plurality of node clusters that complies with the at least one performance rule for supporting the service; and

allocating the service to one of the node clusters that complies with the at least one performance rule, wherein showing each node cluster of the plurality of node clusters that complies with the at least one performance rule for supporting the service, comprises showing a map of each node cluster of the plurality of node clusters that complies with the at least one performance rule for supporting the service, and wherein allocating the service to one of the node clusters that complies with the at least one performance rule comprises allocating the service to one of the node clusters that complies with the at least one performance rule based on the map.

2. The method of claim 1, wherein the showing includes showing at least one server within a first node cluster of said plurality of node clusters that complies with the at least one performance rule for supporting the service.

3. The method of claim 2, further comprising:

allocating the service to a first server of said at least one server.

4. The method of claim 1, wherein collecting the set of performance data representative of the set of physical characteristics of the distributed data processing system comprises probing the distributed data processing system for a round trip time.

5. The method of claim 1, wherein collecting the set of performance data representative of the set of physical characteristics of the distributed data processing system comprises probing the distributed data processing system for a hop count.

6. The method of claim 1, wherein collecting the set of performance data representative of the set of physical characteristics of the distributed data processing system comprises probing the distributed data processing system for a bottleneck link speed.

7. A method for allocating a service in a distributed data processing system, the method comprising:

receiving logical data associated with a logical configuration of the distributed data processing system, the distributed data processing system including a plurality of clients and a plurality of servers, wherein the plurality of clients and the plurality of servers communicate over the distributed data processing system, and wherein each server of the plurality of servers provides at least one assigned service, and wherein the logical data includes data indicating interconnections of the distributed data processing system;

collecting performance data based on the logical data, the performance data being representative of at least one physical characteristic of the distributed data processing system to form collected performance data;

determining, using the collected performance data, cluster data that identifies each node cluster of a plurality of node clusters within the distributed data processing system;

correlating the determined cluster data with at least one performance rule for supporting the service;

showing each node cluster of the plurality of node clusters that complies with the at least one performance rule based on the correlation; and

allocating the service to a node cluster within the distributed data processing system based on the showing, wherein showing each node cluster of the plurality of node clusters that complies with the at least one performance rule based on the correlation comprises showing a map of each node cluster of the plurality of node clusters that complies with the at least one performance rule based on the correlation, and wherein allocating the service to a node cluster within the distributed data processing system based on the showing comprises allocating the service to the node cluster within the distributed data processing system that complies with the at least one performance rule based on the map.

8. The method of claim 7 wherein the performance data includes round trip time, hop count and bottleneck speed.

9. The method of claim 7 wherein the cluster data is determined responsive to a self organizing feature map neural network output.

10. The method of claim 7 further comprising issuing a probe from a module, wherein the logical data is received at the module responsive to issuing the probe, and wherein an engine correlates the determined cluster data with the at least one performance rule.